visualization of chemical reaction dynamics: toward ...€¦ · keywords: molecular beam, chemical...

Review

Visualization of chemical reaction dynamics:Toward understanding complex polyatomic reactions

By Toshinori SUZUKI*1,*2,†

(Communicated by Hiroo INOKUCHI, M.J.A.)

Abstract: Polyatomic molecules have several electronic states that have similar energies.Consequently, their chemical dynamics often involve nonadiabatic transitions between multiplepotential energy surfaces. Elucidating the complex reactions of polyatomic molecules is one of themost important tasks of theoretical and experimental studies of chemical dynamics. This paperdescribes our recent experimental studies of the multidimensional multisurface dynamics ofpolyatomic molecules based on two-dimensional ion/electron imaging. It also discusses ultrafastphotoelectron spectroscopy of liquids for elucidating nonadiabatic electronic dynamics in aqueoussolutions.

Keywords: molecular beam, chemical reaction, nonadiabatic transition, imaging, scattering,photoelectron spectroscopy

1. Introduction: imaging reactivemolecular scattering

Liquid is the most important phase of matter inchemistry and biology. However, it is very difficult tostudy bimolecular reaction dynamics in liquids dueto the billions of collisions occurring per second inliquids. These collisions obscure the initial conditionsof the reacting pair and obliterate the nascentproduct state distributions. Consequently, the mostdetailed studies of bimolecular reactions employcrossed atomic and molecular beams in vacuum.The crossed-beam technique prepares well-definedquantum states and velocities of reactants andinterrogates the products and their velocities. Thismethod was first demonstrated in 1955.1) It wasinitially applied only to the reactions of alkali metalatoms due to its low detection sensitivity of products;this period of research is often referred to as the“alkali age” of molecular beam research. In thelate 1960s, a universal detector was developed. Thisdetector employed electron bombardment ionization

and quadrupole mass analysis,2) and it enabled thevelocity distribution (i.e., speed and angular distri-butions) of virtually any chemical species to bemeasured with mass selection. Crossed-beam experi-ments using this universal detector revealed funda-mental aspects of chemical reaction dynamics forwhich the Nobel Prize for Chemistry was awardedto Lee and Herschbach (with Polanyi) in 1986.However, the universal detector is unable to selectthe quantum (rotational, vibrational and electronic)states of reaction products, which limits the useful-ness of this methodology. Since the 1970s, effortshave been made to develop new experimentaltechniques to perform quantum-state-selective meas-urements of the velocity distributions of reactionproducts.3)–8) Of these techniques, two-dimensionalion imaging is the most powerful.9),10)

Experimental and theoretical studies since the1950s have greatly advanced our understanding ofatom D diatom reactions such as H D H2 andF D H2.11)–18) Extension of such studies to atom D

polyatom and diatom D diatom systems is critical forelucidating the multidimensional dynamics of poly-atomic molecules. When studying these reactions, itis essential to measure the velocity distributions ofquantum-state-selected reaction products. To under-stand why, we compare the atom D diatom reaction,O(1D2) D H2 ! OH D H and the atom D polyatom

*1 Department of Chemistry, Graduate School of Science,Kyoto University, Kyoto, Japan.

*2 RIKEN, Saitama, Japan.† Correspondence should be addressed: T. Suzuki, Depart-

ment of Chemistry, Graduate School of Science, Kyoto University,Kyoto 606-8502, Japan (e-mail: [email protected]).

Proc. Jpn. Acad., Ser. B 89 (2013)No. 1] 1

doi: 10.2183/pjab.89.1©2013 The Japan Academy

http://dx.doi.org/10.2183/pjab.89.1

reaction, O(1D2) D CH4 ! OH D CH3. If the O(1D2)and H2 beams collide with a well-defined kineticenergy, the velocity distribution of H can be usedto determine the internal energy distribution of OHby applying the energy and momentum conservationlaws.19)–22) However, for the reaction O(1D2) D CH4,mass-selected velocity measurements of OH or CH3

do not provide the internal energies of the productsOH or CH3 because their velocities are determinedby the total internal energies of both OH and CH3.

Figure 1 presents the velocity distribution ofthe OH product of the reaction O(1D2) D CH4 !OH D CH3 measured using the universal detector.23)

The anisotropic angular distribution indicates thatthe reaction does not generate any long-lived reactionintermediate. On the other hand, this structurelessspeed distribution is not very useful for elucidatingthe dynamics. The following explanation describeshow quantum-state-selected measurements of prod-ucts provide deeper insights into the dynamics of thereaction O(1D2) D CH4.

The O(1D2) state is a highly reactive metastablestate of an O atom. Since O(1D2) is the primaryphotodissociation product of stratospheric ozone,O3 D h8 ! O(1D2) D O2(1"g), the O(1D2) reaction isof great interest for atmospheric chemistry. O(1D2)reactions with an aliphatic hydrocarbon are repre-sentative “insertion” reactions as they proceed via a

complex produced by inserting O(1D2) into a chemi-cal bond. The high reactivity of O(1D2) originatesfrom the attractive long-range force in the insertionpathway: in the case of O(1D2) D CH4, O(1D2) canattack the C–H bond of methane to form a methanolintermediate that subsequently breaks up intoCH3 D OH (the insertion pathway). On the otherhand, product analyses performed in the 1960s and1970s suggest the existence of a direct “abstraction”pathway:24),25) O(1D2) attacks the H atom collinearlywith the C–H bond and abstracts the H atom toform CH3 D OH without any reaction intermediate(the abstraction pathway). The abstraction pathwayhas an activation barrier. The origins of thesetwo pathways can be understood by consideringthe reaction O(1D2) D H2(1’g

D) ! OH(2&) D H(2S).The electronic symmetry of this system is either 1ABor 1ABB depending on the reflection symmetry of theelectronic wave function of O(1D2) relative to themolecular plane defined by the three atoms becausethe electronic wave function of H2(1’g

D) is symmetricfor reflection. Since O(1D2) is a five-fold degeneratestate, five potential energy surfaces are created byinteraction with H2. Of these, the 11AB state iscorrelated with the ground electronic state of H2Oand provides the insertion pathway, while the 11ABBstate leads to the abstraction pathway. A similarsymmetry argument applies to the local symmetry

Fig. 1. Center-of-mass speed and angular distribution of OH produced by O(1D) D CH4 reaction at a collision energy of 6.5 kcal/mol.The distribution was determined by simulation of the experimental data obtained by the universal crossed-beam apparatus.Reproduced with permission from Ref. 23.

T. SUZUKI [Vol. 89,2

of the O, C and H atoms in O(1D2) D CH4(1A1) !OH(2&) D CH3(2A2BB) and it explains the duality ofthe reaction pathway.

In the early 1990s, we used infrared diode laserabsorption spectroscopy to measure the vibrationalstate distribution of CH3 radicals in the reactionO(1D2) D CH4 ! OH D CH3.26) Radicals were ob-served in low vibrational states of the 82 (out-of-plane bending) mode, even though the structure ofthe CH3 moiety changes from pyramidal to planarduring the reaction.26) This result contrasts with theinverted vibrational distribution of OH.27) Thesestudies of the rovibrational state distributions of theproducts could not identify the elusive abstractionpathway.26),27)

The scattering distributions (especially thebackward scattering distribution) of the productsare critical for identifying the abstraction pathway.Since the insertion pathway has long-range attractiveinteractions between O(1D2) and CH4 in the entrancechannel, collisions with large impact parameters (i.e.,glancing collisions) should dominate the reaction.Consequently, the CH3 products of the insertionpathway are strongly scattered in the forwarddirection (i.e., the initial beam direction of CH4). Ifthe CH3OH complex has a sufficiently long lifetime,it will rotate before breaking up into CH3 and OH,so that the angular distribution of OH will extendin the backward direction.28) On the other hand,the abstraction pathway has an extremely shortreaction time so that the product angular distribu-tion is predominantly determined by the transitionstate geometry. The transition state has a nearlycollinear O–H–C geometry, which provides back-ward scattering. The angular distribution in Fig. 1exhibits strong forward scattering, indicating thatinsertion is the main pathway and that the methanolcomplex has a very short lifetime. The distributionreveals the backward scattering component, but itis not clear whether the back-scattered productsoriginate from the “osculating” (short-lived) CH3OHcomplex28) in the insertion pathway or from anabstraction pathway.

In order to investigate O(1D2) reaction withmethane in greater detail, we constructed a crossedmolecular beam apparatus with two-dimensionalion imaging, as schematically shown in Fig. 2.29)–32)

The key component of this apparatus is the intenseO(1D2) beam source that employs 157 nm photo-dissociation of O2 [O2 D h8 ! O(1D2) D O(3PJ)]. O2

is dissociated in the supersonic expansion of thecarrier (rare) gas to maximize the flux of O(1D2) and

control their velocities; since the speed of O(1D2)essentially becomes the same as that of the carriergas, the collision energy between O(1D2) andmethane can be varied by changing the rare gas(He, Ne, Ar, or Kr) used as the carrier. The O(1D2)atomic beam is crossed with a CD4 molecular beam ina scattering chamber33),34) and the CD3 product isstate-selectively ionized by resonance-enhanced mul-tiphoton ionization via the 3pz Rydberg state using atunable nanosecond UV laser. (CD4 is used becauseCD3 radicals are less predissociative than CH3 in the3pz Rydberg state. The CD3

D is also experimentallymuch less affected than CH3

D by background signalsfrom hydrocarbon contaminants from sources such aspump oils.) The three-dimensional distribution ofCD3

D ions is accelerated in an electric field andprojected onto a two-dimensional position-sensitivedetector.33),34) The observed image is corrected forthe apparatus function and numerically inverted tothe speed and angular distributions of CD3.

Figure 3 shows examples of the speed andangular distributions inverted from the observedimages of the CD3 product in the vibrational groundstate.33) Figures 3(a)–(c) correspond to the finalrotational states of J F 3, 5 and 7, respectively. Allthese distributions exhibit strong forward scattering,in agreement with Fig. 1. The striking feature of thedistributions in Fig. 3 is that the backward scatteringcomponent exhibits a discrete speed distribution. Thekinetic energy of the state-selected CD3 is related tothe internal energy of OD by the following energyconservation relation:

Fig. 2. Schematic illustration of crossed molecular beam ionimaging experiment for studying the O(1D2) D CD4 ! OD D

CD3 reaction.

Visualization of chemical reaction dynamics: Toward understanding complex polyatomic reactionsNo. 1] 3

�H þ Ecoll � ErovibðCD3Þ � Etrans ¼ ErovibðODÞ ½1�where "H is the exothermicity of the reaction, Ecoll isthe collision energy, Erovib(CD3) is the rovibrationalenergy of CD3, Etrans is the total translational energyof both products, and Erovib(OD) is the rovibrationalenergy of OD. The observed speed distributions ofCD3 reveal that the counterpart OD radicals areformed in the vibrationally excited states of v F 4–6.The strong vibrational excitation of OD is typicalfor an abstraction reaction with an early barrier.The discrete ring structure indicates that rotationof OD is only weakly excited. This weak rotationalexcitation is indicative of the collinear transitionstate that is expected for the abstraction pathway;the observed feature of backward scattering is rathersimilar to that of F(2P) D CD4.35) The abstraction/insertion ratio diminishes for higher rotational statesin Fig. 3 because the abstraction process has weakerrotational excitation than the insertion process.The forward scattering is structureless, even in ourstate-resolved measurements. This implies that therotational excitation of the OD counterpart is quitestrong in the insertion mechanism, in agreement withlaser-induced fluorescence measurements of OH forO(1D) D CH4.27)

The angular distributions of CD3 in the excitedvibrational states of 82 (out-of-plane bending) mode(data not shown) exhibit essentially the samecharacteristics as the vibrational ground state.However, as shown in Fig. 4, the CD3 products inthe excited state of a symmetric C–D stretchingvibration (v1) do not exhibit a discrete speeddistribution in the backward direction, indicatingthat CD3 (v1 F 1) is produced only in the insertionpathway; the excitation of C–D stretching is ascribedto vibrational coupling in the methanol intermediate.

The strong forward scattering in Fig. 4 indicates thatthe CD3OD complex has an extremely short lifetime.The vibrational distribution of CD3 measured by thecrossed-beam experiment is consistent with weakvibrational excitation of CH3 observed by infrareddiode laser spectroscopy.26) We measured the rota-tional distribution of CD3 produced by crossed-beamscattering and found that CD3 radicals are createdin low K states, where K is the quantum number ofthe angular momentum projected on the top axisof the CD3 radical. Low K states correspond totumbling rotation of CD3.33) Very recently, weperformed a similar crossed-beam imaging experi-ment on O(1D2) D CH4 ! OH D CH3 and observed

Fig. 3. False-color map of scattering distribution of the CD3 product in the vibrational ground state. The rotational states of theproducts are (A) J F 3, (B) 5, and (C) 7. The left and right sides of each image respectively show the observed image and the two-dimensional slice of three-dimensional scattering distribution. The yellow arrows indicate the directions of the atomic and molecularbeams. Forward scattering means ejection of a CD3 radical in the direction of the CD4 beam. Reproduced with permission fromRef. 33.

Fig. 4. Scattering distribution of the CD3 (v1 F 1) products. TheQ-branch peak of the 111 vibronic band was employed in thismeasurement and the off-resonance image was subtracted. Theleft and right sides of each image respectively show the observedimage and a two-dimensional slice of the three-dimensionalscattering distribution. Comparison with Fig. 3 reveals that thebackward scattering component is missing. Reproduced withpermission from Ref. 33.


essentially the same characteristics as O(1D2) DCD4 ! OD D CD3. Figure 5 shows a graphical de-piction of the reaction mechanism of O(1D2) DCH4 ! OH D CH3.

Theoretical estimates of the reaction barrierheight have ambiguities of ca. 2 kcal/mol. Ab initiocalculations using CASPT2 and the basis set 6-311G(2df, 2pd) predict that the insertion pathway isbarrierless, whereas the abstraction pathway has abarrier height of 1.2 kcal/mol.36) Restricted internallycontracted multireference configuration interactioncalculations at the complete basis set limit predictthat the insertion pathway will have a barrier heightof 1.47 kcal/mol (without correcting for the zero-point energy).37) Further experimental studies arecurrently being conducted in our laboratory toinvestigate the reaction barriers and the collisionenergy dependence of the dynamics. Accurate dy-namical calculations of the ab initio potential energysurfaces are highly computationally demanding, yetthey will be very useful for understanding thedynamics.

2. Ultrafast photoelectron imaging: real-timeobservation of nonadiabatic dynamics

Although sophisticated crossed-beam ion imag-ing experiments provide fine details of reactiondynamics, they observe only the products after areaction occurs. Some important aspects of chemicalreaction dynamics can be studied only by real-time

observations. One example is the nonadiabaticdynamics of polyatomic molecules. The Born–Oppenheimer approximation facilitates a quantum-mechanical description of a chemical reaction38) byenabling the equations of motions for electrons andnuclei to be separated. In the Born–Oppenheimerapproximation, chemical reactions are viewed asclassical trajectories or quantum-mechanical wavepacket motions of the nuclear geometry on electronicpotential energy surfaces. However, the Born–Oppenheimer approximation breaks down whendifferent electronic states are close in energy; thisgives rise to complex yet interesting nonadiabaticdynamics. Nonadiabatic transitions frequently occurin polyatomic molecules because they have severalelectronic states with similar energies and theirmultidimensional potential energy surfaces fre-quently undergo avoided crossings. It is thus im-portant to identify nonadiabatic transitions inexperimental and theoretical studies of polyatomicreactions.

Real-time observations of bimolecular reactionsare not currently feasible because the timing ofcollisions cannot be controlled more accurately thanthe reaction time. Consequently, bimolecular dynam-ics is studied only from the asymptotic states ofproducts using the crossed molecular beam scatter-ing, as described in the previous section. On the otherhand, ultrafast spectroscopy permits real-time ob-servation of the nonadiabatic dynamics of photo-induced unimolecular reactions.

Pyrazine (C4H4N2) is one of the most funda-mental heteroaromatic molecules.39)–49) Its broadS2(1B2u, ::*)–S0(1Ag) photoabsorption spectrumsuggests that the S2 state undergoes ultrafastelectronic deactivation (internal conversion) within30 fs. This ultrafast deactivation originates from theconical intersection between the S2 and S1(1B3u, n:*)potential energy surfaces (see Fig. 6).41) In a diatomicmolecule, the von Neumann–Wigner noncrossing ruleprohibits degeneracy of electronic states with thesame symmetry so that their potential energy curveshave avoided crossings.50) For a nonlinear polyatomicmolecule with N vibrational degrees of freedom,potential energy surfaces can be degenerate to form aseam of crossings in N–2 dimensional space (knownas a conical intersection). The conical intersection isthe most important topographical feature of multi-dimensional surfaces of polyatomic molecules.51)–55)

Ultrafast internal conversion of pyrazine via conicalintersection has been extensively investigated theo-retically.39)–49)

Fig. 5. Graphical presentation of the two reaction pathways.Insertion occurs with a large impact parameter and results inforward scattering of rotationally excited products. Abstractionreaction occurs via a collinear transition state and results in thebackward scattering of weakly excited products.


Since a nonadiabatic transition abruptly altersthe electronic character, it is most sensitivelydetected by measuring the electron distribution.However, the total electron density of a molecule isnot the most suitable observable for this purposebecause the total electron density distribution ina molecule does not change at the instant of anonadiabatic transition. Photoelectron spectroscopyenables selective observation of valence electrons thatplay central roles in nonadiabatic transitions. Theangular distribution of a photoelectron is a partic-ularly important observable because a nonadiabatictransition changes the three-dimensional shape of anoccupied electron orbital and affects the photoelec-tron wavefunction. Ultrafast photoelectron imagingvisualizes the photoelectron speed and angulardistributions as a movie with a femtosecond timeresolution and enables identification of nonadiabatictransitions.56)–58)

How are the photoelectron kinetic energy andangular distributions expressed? In the absence of anexternal electromagnetic field, molecules in the gasphase have random alignment of the molecular axes.One-photon absorption induced by a linearly polar-ized pump pulse promotes molecules to an excitedstate whose transition dipole moments are favorablyaligned with the electric field vector of the pumppulse. The probe pulse induces ejection of a photo-electron from this aligned ensemble of molecules inthe excited state. Based on these two optical steps,the final photoelectron kinetic energy and angulardistribution can be expressed by

IðE; �; tÞ ¼ �ðE; tÞ4�

f1þ �2ðE; tÞP2ðcos �Þþ �4ðE; tÞP4ðcos �Þg ½2�

for a probe pulse whose linear polarization is parallelto that of the pump pulse. In this expression, E isthe photoelectron kinetic energy, 3 is the electronejection angle relative to the laser polarizationdirection, and t is the pump–probe time delay.Pn(x) is the nth-order Legendre polynomial. <(E, t)represents the photoelectron kinetic energy distribu-tion (i.e., the photoelectron spectrum). The aniso-tropy parameters, O2(E, t) and O4(E, t), containinformation about the spatial distribution of anionized orbital. Two-dimensional maps of <(E, t)and O2(E, t) against time and energy are the keyobservables for analyzing nonadiabatic dynamics.

Ultrafast photoelectron imaging combines theultrafast pump–probe method and two-dimensionalposition-sensitive detection of electrons. An expand-ing distribution of photoelectrons is created when alaser pulse ionizes molecules in the vacuum chamber.Photoelectron imaging accelerates these electronsin a static electric field and projects them onto atwo-dimensional position-sensitive detector (Fig. 7).Since this method enables the total speed andangular distributions of photoelectrons to be simul-taneously measured on a shot-to-shot basis, itachieves unprecedentedly high efficiency and accu-racy. Ultrashort UV pulses required for this experi-ment can be generated by nonlinear optical processesin rare gases. Since a gas has a small nonlinearsusceptibility, a long interaction length between laserpulses and a gas medium is required to realizeefficient wavelength conversion. To obtain a suffi-ciently long interaction length, we employ filamenta-tion propagation of intense laser pulses; when anintense laser pulse propagates through a gas, it is self-focused by the optical Kerr effect and defocusedby the plasma created by ionization of the gas.Consequently, a laser beam propagates through a gaswhile maintaining a small beam diameter for a lengththat is considerably greater than the confocal pa-rameter.59) We introduce 775 and 388 nm pulses toNe (0.08MPa) and induce cascaded four-wave mixingof 2!!388 � !!775 ! !!264 and !!264 þ !!388 �!!775 ! !!198 to generate 264 and 198 nm pulses(Fig. 7).60) These pulses are separated and com-pressed using a grating-based compressor to achievepulse durations of 14 and 17 fs, respectively.60),61)

We excited pyrazine molecules in a supersonicmolecular beam using the 264 nm pulse and sub-

Fig. 6. Conical intersection of S2 and S1 adiabatic potentialenergy surfaces of pyrazine in the two-dimensional spacespanned by Q10a and Q6a. Reproduced with permission fromRef. 47.


sequently ionized them with the 198 nm pulse.Figure 8b shows the photoelectron intensity observedas a function of the pump–probe time delay. Theintensity decays rapidly in the first 100 fs andsubsequently plateaus.62),63) This time profile iswell explained by assuming rapid decay of opticallyexcited S2 (red) and formation of S1 (blue) byinternal conversion from S2. In the negative delaytime range, the 198 nm pulse excites pyrazine to S3and the 264 nm pulse ionizes (green); since this isunrelated to the present discussion, we neglect ithere. The S2!S1 internal conversion time constantsare determined as 23 ’ 4 fs for pyrazine-h4 and20 ’ 2 fs for pyrazine-d4. Close examination revealsthat the signal oscillates in the plateau region. TheFourier transform of this feature (t > 50 fs) exhibits afrequency component of 560 ’ 40 cm!1 for pyrazineand 550 ’ 40 cm!1 for pyrazine-d4. Based on thesefrequencies, the oscillation is assigned to vibrationalwave packet motion of mode 6a in S1 created byinternal conversion.

We now examine the kinetic energy distributionof photoelectrons. Figure 8a shows a time–energymap of <(E, t). The colors yellow, red and violetindicate decreasing intensities. <(E, t) exhibits nomarked change in distribution; the intensity dimin-ishes gradually from a low to high kinetic energy atall times. The kinetic energy distribution does notchange on internal conversion, because photoioniza-tion predominantly occurs as D0(n!1)AS1(n,:*) andD1(:!1)AS2(:,:*) (see Fig. 9) and these verticalionization energies are almost identical.64),65) In otherwords, removal of a :* electron always requires thesame energy.

How is the photoelectron angular distribution?Figure 8c shows a time–energy map of O2(E, t);the positive (blue-green) and negative (red) valuesrespectively indicate preferential ejection of anelectron parallel and perpendicular to the probe laserpolarization. The most distinctive feature in O2(E, t)is the sudden change at ca. 30 fs; this notablechange in O2(E, t) is unambiguous evidence forS2!S1 internal conversion. (The influence of molecu-lar rotation can be neglected on this time scale.)After 30 fs, O2(E, t) does not change, because theelectronic character remains (n,:*). The lack ofrestoration of the (:,:*) character is related to thephotoexcitation energy; in our experiment, photo-excitation of pyrazine near the S2 origin preparesa wave packet with a small vibrational excess energyin S2. Consequently, if the vibrational energy flowsinto various modes in S1, the wave packet has noopportunity to return to the Franck–Condon regionin S2.

Ultrafast S2!S1 internal conversion of pyrazinehas been considered as a two-state problem via aconical intersection. The above-described results ofultrafast photoelectron imaging are consistent withthis well-accepted picture. On the other hand, recentquantum chemical calculations suggest that opticallydark electronic states of 1Au and 1B2g are alsoinvolved in the ultrafast internal conversion dynam-ics of pyrazine and that a fraction of the populationin S2 flows into these states.40),66) The proposedmechanism is currently being investigated in ourlaboratory by ultrafast photoelectron imaging using avacuum UV probe pulse (160 nm).

Fig. 7. Schematic diagram of filamentation four-wave mixing to generate sub-20 fs deep UV pulses and our photoelectron imaging setup.


3. New challenge: ultrafast photoelectronspectroscopy of nonadiabatic

dynamics in liquids

The preceding two sections discussed chemicaldynamics in the gas phase. The study of gas-phasereactions provides the basis for understanding morecomplex dynamics in condensed phases. The greatestdifference between gas and condensed phases is thatsolvent effects play important roles in the latterphases. Despite not fully understanding the mecha-nisms involved, chemists have utilized solvent effectsfor centuries to alter potential energy surfacesand reaction dynamics. It is thus exciting to applyultrafast photoelectron spectroscopy to liquids toinvestigate static and dynamic solvent effects

and nonadiabatic processes, particularly in aqueoussolutions.

Siegbahn67) pioneered photoelectron spectrosco-py of liquids in the 1970s. The technical problem tointroduce volatile liquids into a photoelectron spec-trometer was solved to a large extent by a liquidbeam technique developed by Faubel and co-workers68) in the 1980s. Winter and coworkers69)

performed photoelectron spectroscopy of liquidbeams with third-generation synchrotron radiation.Kondow and Mafune combined a laser with liquidbeams to perform photoionization mass spectrometryin 1992.70) It took nearly two decades before ultrafastphotoelectron spectroscopy of liquids was realizedin 2010.71)–73) This section describes this very newresearch field.

An electron passing through a liquid is scatteredby solvent molecules. This scattering determinesthe escape depth of a photoelectron (and hence theprobing depth of photoelectron spectroscopy). There-fore, the inelastic mean free path of an electron inthe bulk is a very important quantity. However, theinelastic mean free path is not well understood forany liquid. It is thus necessary to speculate about itbased on those in solids. Experimental studies of solidmaterials show that the mean free path becomesminimal (<1 nm) for kinetic energies in the range50–100 eV.74) Inelastic scattering of an electron inamorphous ice indicates that electron energy lossesgreater than 1 eV rarely occurs for kinetic energiessmaller than the HOMO–LUMO gap (97 eV) of bulkwater.75) This is because such low-energy electronscannot excite the valence electrons of water and

Fig. 8. (a) Time-evolution of photoelectron kinetic energydistribution, <(E, t). (b) Temporal profiles of total photoelectronsignals in (1 D 1B) TRPEI of pyrazine-h4. The observed data arewell explained by three components: single-exponential decay ofS2 (red); corresponding increase in S1 (blue) in the positive timedelay; and single-exponential decay of S3 (green) in the negative-time delay. The fitting result is shown by the solid line. (c) Timeevolution of photoelectron angular anisotropy parameterO2(E, t).

Fig. 9. Photoionization schemes of pyrazine within the frame-work of the frozen-core approximation.


hence interact only with phonons. Since electron–phonon scattering has a considerably smaller crosssection than electron–electron scattering, the inelas-tic mean free path (and hence the probing depth ofphotoelectron spectroscopy) increases with decreas-ing kinetic energy. (Another extreme case is hard-X-ray photoelectron spectroscopy, which is also bulksensitive.)

Using ultrafast photoelectron spectroscopy ofliquids, we studied the charge-transfer-to-solvent(CTTS) reaction from I!(aq) to bulk water (seeFig. 10).76)–78) A hydrated I! atom has an electronbinding energy of 8.0 eV and its ground state lies inthe band gap of bulk water; the electron bindingenergy of bulk water is 11.2 eV.79) The UV absorptionmaximum of I!(aq) is at 225 nm, corresponding toexcitation to the metastable (CTTS) electronic state.The CTTS state rapidly undergoes exothermicelectron transfer to bulk water. When the excesselectron migrates into bulk water, a local hydrogen-bonding network around the I atom and the excesselectron change drastically due to reorientation ofwater molecules; I! and an excess electron cloud arehydrophilic, whereas a neutral I atom is hydrophobic.Thus, the CTTS dynamics is strongly coupled withthe solvation dynamics. The I atom has no internaldegrees of freedom besides electronic motion, makingit a suitable solute for investigating the CTTSdynamics. We excited I!(aq) with the 226 nm pumppulse and photodetached the excess electron witha time-delayed 260 nm probe pulse. A hemisphericalelectron energy analyzer with extensive differentialpumping was employed.80)

Figure 11 compares the photoelectron signalintensity, I(t), observed for aqueous NaI solutions ofH2O and D2O as a function of the pump–probe delaytime. The figure indicates a clear isotope effect: thephotoelectron intensity in the first few picosecondsdiminishes much faster in D2O than in H2O, while theopposite is true in the subsequent 30 ps. Anotherisotope effect is seen in Fig. 12 that shows <(E, t) forsolutions in H2O and D2O. <(E, t) has similar generalfeatures for H2O and D2O; however, the spectralwidth of <(E, t) is narrower and its intensitydiminishes faster in D2O than in H2O. The timedependences of the photoelectron intensity andkinetic energy distribution can be explained by themechanism illustrated in Fig. 13. After photoexcita-tion of I! to a metastable state, the excess electrondistribution becomes asymmetric in the hydrationshell and separated from a neutral I atom. Calcu-lations for Cl! and water suggest that attractiveinteractions occur between a neutral atom and ahydrated electron;81) a similar attraction is antici-pated between an I atom and a hydrated electron.The hydrated electron and I atom are finallyseparated by diffusion. The photon energy of ourprobe laser pulse is higher than the binding energiesof the CTTS state, the contact pair state, the solventseparated state, and a free hydrated electron, but it islower than the ground state of I!(aq). Therefore, the

Fig. 10. Schematic energy diagram of time-resolved photoelec-tron spectroscopy of charge-transfer-to solvent reaction. Repro-duced with permission from Ref. 94.

Fig. 11. Photoelectron signal intensity as a function of pump–probe time delay observed for charge-transfer-to-solvent reactionfrom I! to bulk water: the sample solutions were 0.1M aqueousNaI solution in H2O (black) and D2O (blue). The schematicenergy diagram of the experiment is shown in Fig. 10. The crosscorrelation of the pump and probe laser pulses was ca. 300 fs.A hemispherical electron energy analyzer was used to measurethe photoelectron spectrum at each time delay and the spectrumwas integrated over the photoelectron kinetic energy to obtaineach data point. From Ref. 94.


disappearance of the photoelectron signal in Fig. 11is ascribed to internal conversion to the groundelectronic state of I!(aq); i.e., geminate recombina-tion of an electron and an I atom. In the initial steps

of the CTTS reaction, H2O has a faster librationalresponse than D2O. Consequently, the quantum yieldof internal conversion and the loss of electron signalare greater in D2O during the early stages of the

Fig. 12. Two-dimensional false-color map for photoelectron kinetic energy distributions measured for charge-transfer-to-solvent reactionfrom I! to bulk water for various pump–probe time delays. The schematic energy diagram of the experiment is shown in Fig. 9.Experimental data for 0.1M aqueous NaI solution in (a) H2O and (b) D2O. (c) and (d) show the results of global fitting ofexperimental data using a kinetic model. From Ref. 94.

Fig. 13. Graphical presentation of our kinetic model with some representative time constants for the charge-transfer-to-solvent reactionfrom I! to bulk water.


reaction. On the other hand, internal conversion isnearly two times faster in H2O than in D2O duringthe later stages of the CTTS reaction.

The CTTS reaction for bulk water ultimatelycreates hydrated electrons,56),71)–73),82)–94) which arethe most important transient species in radiationchemistry and biology. The electron binding energyof a hydrated electron has been speculated to be3.3 eV, although no direct measurement has beenperformed. Our photoelectron spectroscopy of liquidshas determined that the electron binding energy of ahydrated electron is 3.4 eV and that the energies ofsolvated electrons in methanol and ethanol are 3.5and 3.3 eV, respectively.66),72),73),94),95) Interestingly,the electron binding energy of a hydrated electronagrees with a previous estimate (3.3 eV) obtainedby extrapolating the binding energies in negativelycharged water clusters.84) On the other hand, theelectron binding energy of a solvated electron inmethanol is much greater than the estimate(2.6 eV) from corresponding cluster values.96) Thecause of this interesting difference is currently beinginvestigated.

Ultrafast photoelectron spectroscopy of liquids isa very young field and there are many unsolvedproblems for this methodology. On the other hand,it is promising for elucidating electronic dynamicsin solution and providing important new knowledgefor chemistry, physics and biology.

Acknowledgments

I would like to thank H. Kohguchi, Y. Ogi, Y.Suzuki, T. Fuji and T. Horio for their contributionsto the research described in this paper.

References

1) Taylor, E.H. and Datz, S. (1955) Study of chemicalreaction mechanisms with molecular beams—theReaction of K with HBr. J. Chem. Phys. 23, 1711–1718.

2) Lee, Y.T., McDonald, J.D., LeBreton, P.R. andHerschbach, D.R. (1969) Molecular beam reactivescattering apparatus with electron bombardmentdetector. Rev. Sci. Instrum. 40, 1402–1408.

3) Kinsey, J.L. (1977) Fourier-transform doppler spec-troscopy—New means of obtaining velocity-angledistributions in scattering experiments. J. Chem.Phys. 66, 2560–2565.

4) Murphy, E.J., Brophy, J.H., Arnold, G.S., Dimpfl,W.L. and Kinsey, J.L. (1979) Velocity andangular-distributions of reactive collisions fromFourier-transform Doppler spectroscopy—1st ex-perimental results. J. Chem. Phys. 70, 5910–5911.

5) Buntine, M.A., Baldwin, D.P., Zare, R.N. and

Chandler, D.W. (1991) Application of ion imagingto the atom-molecule exchange-reaction—HDHI!H2DI. J. Chem. Phys. 94, 4672–4675.

6) Girard, B., Billy, N., Gouédard, G. and Vigué, J.(1991) Differential cross-section as a function ofthe product internal state—the FDI2!IF(v,J)DIreactive collision. J. Chem. Phys. 95, 4056–4069.

7) Suits, A.G., Bontuyan, L.S., Houston, P.L. andWhitaker, B.J. (1992) Differential cross-sectionsfor state-selected products by direct imaging—ArDNO. J. Chem. Phys. 96, 8618–8620.

8) Schnieder, L., Seekamprahn, K., Borkowski, J.,Wrede, E., Welge, K.H., Aoiz, F.J., Banares, L.,Dmello, M.J., Herrero, V.J., Rabanos, V.S. andWyatt, R.E. (1995) Experimental studies andtheoretical predictions for the HDD2!HDDDreaction. Science 269, 207–210.

9) Chandler, D.W. and Houston, P.L. (1987) Two-dimensional imaging of state-selected photodisso-ciation products detected by multiphoton ioniza-tion. J. Chem. Phys. 87, 1445–1447.

10) Eppink, A.T.J.B. and Parker, D.H. (1997) Velocitymap imaging of ions and electrons using electro-static lenses: Application in photoelectron andphotofragment ion imaging of molecular oxygen.Rev. Sci. Instrum. 68, 3477–3484.

11) Continetti, R.E., Balko, B.A. and Lee, Y.T. (1990)Crossed molecular-beams study of the reactionDDH2!DHDH at collision energies of 0.53 and1.01 eV. J. Chem. Phys. 93, 5719–5740.

12) Wrede, E., Schnieder, L., Welge, K.H., Aoiz, F.J.,Banares, L., Herrero, V.J., MartinezHaya, B. andRabanos, V.S. (1997) The HDD2 reaction in thevicinity of the conical intersection. J. Chem. Phys.106, 7862–7864.

13) Schnieder, L., SeekampRahn, K., Wrede, E. andWelge, K.H. (1997) Experimental determination ofquantum state resolved differential cross sectionsfor the hydrogen exchange reaction HDD2!HDDD. J. Chem. Phys. 107, 6175–6195.

14) Neumark, D.M., Wodtke, A.M., Robinson, G.N.,Hayden, C.C. and Lee, Y.T. (1985) Molecular-beam studies of the FDH2 reaction. J. Chem. Phys.82, 3045–3066.

15) Neumark, D.M., Wodtke, A.M., Robinson, G.N.,Hayden, C.C., Shobatake, K., Sparks, R.K.,Schafer, T.P. and Lee, Y.T. (1985) Molecular-beam studies of the FDD2 and FDHD reactions. J.Chem. Phys. 82, 3067–3077.

16) Skodje, R.T., Skouteris, D., Manolopoulos, D.E.,Lee, S.H., Dong, F. and Liu, K.P. (2000)Resonance-mediated chemical reaction: FDHD!HFDD. Phys. Rev. Lett. 85, 1206–1209.

17) Qiu, M.H., Ren, Z.F., Che, L., Dai, D.X., Harich,S.A., Wang, X.Y., Yang, X.M., Xu, C.X., Xie,D.Q., Gustafsson, M., Skodje, R.T., Sun, Z.G. andZhang, D.H. (2006) Observation of Feshbachresonances in the FDH2!HFDH reaction. Science311, 1440–1443.

18) Che, L., Ren, Z.F., Wang, X.G., Dong, W.R., Dai,D.X., Wang, X.Y., Zhang, D.H., Yang, X.M.,Sheng, L.S., Li, G.L., Werner, H.J., Lique, F. and


Alexander, M.H. (2007) Breakdown of the Born-Oppenheimer approximation in the FDo-D2!DFDD reaction. Science 317, 1061–1064.

19) Hsu, Y.T., Wang, J.H. and Liu, K.P. (1997)Reaction dynamics of O(1D)DH2, D2, and HD:Direct evidence for the elusive abstraction path-way and the estimation of its branching. J. Chem.Phys. 107, 2351–2356.

20) Hsu, Y.T. and Liu, K.P. (1997) Product state(s)-resolved differential cross section of the reactionO(1D)DHD!OH(v, j)DD. J. Chem. Phys. 107,1664–1667.

21) Liu, X.H., Lin, J.J., Harich, S., Schatz, G.C. andYang, X.M. (2000) A quantum state-resolvedinsertion reaction: O(1D)DH2(JF0)!OH(2&, v,N)DH(2S). Science 289, 1536–1538.

22) Liu, X.H., Lin, J.J., Harich, S.A. and Yang, X.M.(2001) State-to-state dynamics for O(1D)DD2 !ODDD: Evidence for a collinear abstractionmechanism. Phys. Rev. Lett. 86, 408–411.

23) Lin, J.J., Shu, J., Lee, Y.T. and Yang, X. (2000)Multiple dynamical pathways in the O(1D)DCH4reaction: A comprehensive crossed beam study. J.Chem. Phys. 113, 5287–5301.

24) Paraskevopoulos, G. and Cvetanovic, R.J. (1969)Competitive reactions of excited oxygen atoms,O(1D). J. Am. Chem. Soc. 91, 7572–7577.

25) Lin, C.L. and Demore, W.B. (1973) Reactions ofO(1D) with methane and ethane. J. Phys. Chem.77, 863–869.

26) Suzuki, T. and Hirota, E. (1993) Vibrationaldistribution of CH3 produced by the reaction ofO(1D2) atom with CH4. J. Chem. Phys. 98, 2387–2398.

27) Park, C.R. and Wiesenfeld, J.R. (1991) Full charac-terization of OH product energetics in the reactionof O(1D2) with hydrocarbons. J. Chem. Phys. 95,8166–8177.

28) Miller, W.B., Safron, S.A. and Herschbach, D.R.(1967) Exchange reactions of alkali atoms withalkali halides—A collision complex mechanism.Discuss. Faraday Soc. 44, 108–122.

29) Yonekura, N., Gebauer, C., Kohguchi, H. andSuzuki, T. (1999) A crossed molecular beamapparatus using high-resolution ion imaging. Rev.Sci. Instrum. 70, 3265–3270.

30) Kohguchi, H., Suzuki, T. and Alexander, M.H.(2001) Fully state-resolved differential cross sec-tions for the inelastic scattering of the open-shellNO molecule by Ar. Science 294, 832–834.

31) Kohguchi, H., Suzuki, T., Nanbu, S., Ishida, T.,Mil’nikov, G.V., Oloyede, P. and Nakamura, H.(2008) Collision energy dependence of the O(1D)DHCl!OHDCl(2P) reaction studied by crossedbeam scattering and quasiclassical trajectorycalculations on ab initio potential energy surfaces.J. Phys. Chem. A 112, 818–825.

32) Kohguchi, H. and Suzuki, T. (2006) A crossedmolecular beam imaging study of the O(1D2)DHCl!OHDCl(2PJF3/2,1/2) reaction. ChemPhys-Chem 7, 1250–1257.

33) Kohguchi, H., Ogi, Y. and Suzuki, T. (2011)

Rovibrational state specific scattering distributionsof the O(1D)DCD4!ODDCD3 (v1, v2, N) reaction.Phys. Chem. Chem. Phys. 13, 8371–8378.

34) Kohguchi, H., Ogi, Y. and Suzuki, T. (2008)Reaction mechanism duality in O(1D2)DCD4!ODDCD3 identified from scattering distributionsof rotationally state selected CD3. Phys. Chem.Chem. Phys. 10, 7222–7225.

35) Lin, J.J., Zhou, J.G., Shiu, W.C. and Liu, K.P.(2003) State-specific correlation of coincidentproduct pairs in the FDCD4 reaction. Science300, 966–969.

36) Hernando, J., Millan, J., Sayos, R. and Gonzalez, M.(2003) Ab initio study of the O(1D)DCH4(X 1A1)!OH(X 2&)DCH3(X 2A2BB) reaction: ground andexcited potential energy surfaces. J. Chem. Phys.119, 9504–9512.

37) Yu, H.-G. and Muckerman, J.T. (2004) MRCIcalculations of the lowest potential energy surfacefor CH3OH and direct ab initio dynamics simu-lations of the O(1D) D CH4 reaction. J. Phys.Chem. A 108, 8615–8623.

38) Born, M. and Oppenheimer, R. (1927) Quantumtheory of molecules. Ann. Phys. 84, 0457–0484.

39) Werner, U., Mitric, R. and Bonacic-Koutecky, V.(2010) Simulation of time resolved photoelectronspectra with Stieltjes imaging illustrated on ultra-fast internal conversion in pyrazine. J. Chem.Phys. 132, 174301.

40) Werner, U., Mitric, R., Suzuki, T. and Bonacic-Koutecky, V. (2008) Nonadiabatic dynamics with-in the time dependent density functional theory:Ultrafast photodynamics in pyrazine. Chem. Phys.349, 319–324.

41) Schneider, R. and Domcke, W. (1988) S1-S2 conicalintersection and ultrafast S2!S1 internal-conver-sion in pyrazine. Chem. Phys. Lett. 150, 235–242.

42) Seel, M. and Domcke, W. (1991) Femtosecondtime-resolved ionization spectroscopy of ultrafastinternal-conversion dynamics in polyatomic-molecules—Theory and computational studies. J.Chem. Phys. 95, 7806–7822.

43) Seidner, L., Stock, G., Sobolewski, A.L. and Domcke,W. (1992) Ab initio characterization of the S1-S2conical intersection in pyrazine and calculation ofspectra. J. Chem. Phys. 96, 5298–5309.

44) Sobolewski, A.L., Woywod, C. and Domcke, W.(1993) Ab initio investigation of potential-energysurfaces involved in the photophysics of benzeneand pyrazine. J. Chem. Phys. 98, 5627–5641.

45) Stock, G. and Domcke, W. (1993) Femtosecondspectroscopy of ultrafast nonadiabatic excited-state dynamics on the basis of ab initio potential-energy surfaces—the S2 state of pyrazine. J. Phys.Chem. 97, 12466–12472.

46) Stock, G., Woywod, C., Domcke, W., Swinney, T.and Hudson, B.S. (1995) Resonance raman-spectroscopy of the S1 and S2 states of pyrazine—experiment and first principles calculation ofspectra. J. Chem. Phys. 103, 6851–6860.

47) Woywod, C., Domcke, W., Sobolewski, A.L. andWerner, H.J. (1994) Characterization of the S1-S2


conical intersection in pyrazine using ab initiomulticonfiguration self-consistent-field and multi-reference configuration-interaction methods. J.Chem. Phys. 100, 1400–1413.

48) Raab, A., Worth, G.A., Meyer, H.D. andCederbaum, L.S. (1999) Molecular dynamics ofpyrazine after excitation to the S2 electronic stateusing a realistic 24-mode model Hamiltonian. J.Chem. Phys. 110, 936–946.

49) Worth, G.A., Meyer, H.D. and Cederbaum, L.S.(1996) The effect of a model environment on the S2absorption spectrum of pyrazine: A wave packetstudy treating all 24 vibrational modes. J. Chem.Phys. 105, 4412–4426.

50) von Neumann, J. and Wigner, E. (1929) Concerningthe behaviour of eigenvalues in adiabatic processes.Phys. Z. 30, 467–470.

51) Bernardi, F., Olivucci, M. and Robb, M.A. (1996)Potential energy surface crossings in organicphotochemistry. Chem. Soc. Rev. 25, 321–328.

52) Domcke, W., Yarkony, D.R. and Koppel, H. (2011)Conical Intersections: Theory, Computation andExperiment, Vol. 17, World Scientific PublishingCo. Pte. Ltd., Singapore.

53) Domcke, W., Yarkony, D.R. and Koppel, H. (2004)Conical Intersections: Electronic Structure, Dy-namics & Spectroscopy, Vol. 15, World ScientificPublishing Co. Pte. Ltd., Singapore.

54) Yarkony, D.R. (1998) Conical intersections: Diabol-ical and often misunderstood. Acc. Chem. Res. 31,511–518.

55) Yarkony, D.R. (1996) Diabolical conical intersec-tions. Rev. Mod. Phys. 68, 985–1013.

56) Suzuki, T. and Whitaker, B.J. (2001) Non-adiabaticeffects in chemistry revealed by time-resolvedcharged-particle imaging. Int. Rev. Phys. Chem.20, 313–356.

57) Suzuki, T. (2006) Femtosecond time-resolved photo-electron imaging. Annu. Rev. Phys. Chem. 57,555–592.

58) Suzuki, T. (2012) Time-resolved photoelectron spec-troscopy of non-adiabatic electronic dynamics ingas and liquid phases. Int. Rev. Phys. Chem. 31,265–318.

59) Chin, S.L. (2010), Femtosecond Laser Filamenta-tion. Springer, New York.

60) Fuji, T., Horio, T. and Suzuki, T. (2007) Generationof 12 fs deep-ultraviolet pulses by four-wave mixingthrough filamentation in neon gas. Opt. Lett. 32,2481–2483.

61) Fuji, T., Suzuki, T., Serebryannikov, E.E. andZheltikov, A. (2009) Experimental and theoreticalinvestigation of a multicolor filament. Phys. Rev.A 80, 063822.

62) Wang, L., Kohguchi, H. and Suzuki, T. (1999)Femtosecond time-resolved photoelectron imaging.Faraday Discuss. 113, 37–46.

63) Stert, V., Farmanara, P. and Radloff, W. (2000)Electron configuration changes in excited pyrazinemolecules analyzed by femtosecond time-resolvedphotoelectron spectroscopy. J. Chem. Phys. 112,4460–4464.

64) Oku, M., Hou, Y., Xing, X., Reed, B., Xu, H., Chang,C., Ng, C.-Y., Nishizawa, K., Ohshimo, K. andSuzuki, T. (2008) 3s Rydberg and cationic statesof pyrazine studied by photoelectron spectroscopy.J. Phys. Chem. A 112, 2293–2310.

65) Innes, K.K., Ross, I.G. and Moomaw, W.R. (1988)Electronic states of azabenzenes and azanaphtha-lenes—A revised and extended critical-review. J.Mol. Spectrosc. 132, 492–544.

66) Lin, C.K., Niu, Y.L., Zhu, C.Y., Shuai, Z.G. and Lin,S.H. (2011) The role of the n:* 1Au state in thephotoabsorption and relaxation of pyrazine. Chem.Asian J. 6, 2977–2985.

67) Siegbahn, H. (1985) Electron-spectroscopy forchemical-analysis of liquids and solutions. J. Phys.Chem. 89, 897–909.

68) Faubel, M., Steiner, B. and Toennies, J.P. (1997)Photoelectron spectroscopy of liquid water, somealcohols, and pure nonane in free micro jets. J.Chem. Phys. 106, 9013–9031.

69) Winter, B. and Faubel, M. (2006) Photoemissionfrom liquid aqueous solutions. Chem. Rev. 106,1176–1211.

70) Mafune, F., Takeda, Y., Nagata, T. and Kondow, T.(1992) Formation and ejection of cluster ions froma liquid beam of aniline—ethanol solution by laserphotoionization. Chem. Phys. Lett. 199, 615–620.

71) Tang, Y., Shen, H., Sekiguchi, K., Kurahashi, N.,Mizuno, T., Suzuki, Y.I. and Suzuki, T. (2010)Direct measurement of vertical binding energy ofa hydrated electron. Phys. Chem. Chem. Phys. 12,3653–3655.

72) Siefermann, K.R., Liu, Y.X., Lugovoy, E., Link, O.,Faubel, M., Buck, U., Winter, B. and Abel, B.(2010) Binding energies, lifetimes and implicationsof bulk and interface solvated electrons in water.Nat. Chem. 2, 274–279.

73) Lubcke, A., Buchner, F., Heine, N., Hertel, I.V. andSchultz, T. (2010) Time-resolved photoelectronspectroscopy of solvated electrons in aqueous NaIsolution. Phys. Chem. Chem. Phys. 12, 14629–14634.

74) Seah, M.P. and Dench, W.A. (1979) Quantitativeelectron spectroscopy of surfaces; a standard database for electron inelastic mean free paths in solids.Surf. Interface Anal. 1, 2–11.

75) Michaud, M., Wen, A. and Sanche, L. (2003) Crosssections for low-energy (1–100 eV) electron elasticand inelastic scattering in amorphous ice. Radiat.Res. 159, 3–22.

76) Iglev, H., Trifonov, A., Thaller, A., Buchvarov, I.,Fiebig, T. and Laubereau, A. (2005) Photoioniza-tion dynamics of an aqueous iodide solution: thetemperature dependence. Chem. Phys. Lett. 403,198–204.

77) Kloepfer, J.A., Vilchiz, V.H., Lenchenkov, V.A.,Germaine, A.C. and Bradforth, S.E. (2000) Theejection distribution of solvated electrons gener-ated by the one-photon photodetachment ofaqueous I- and two-photon ionization of thesolvent. J. Chem. Phys. 113, 6288–6307.

78) Chen, X.Y. and Bradforth, S.E. (2008) The ultrafast


dynamics of photodetachment. Annu. Rev. Phys.Chem. 59, 203–231.

79) Winter, B., Weber, R., Hertel, I.V., Faubel, M.,Jungwirth, P., Brown, E.C. and Bradforth, S.E.(2005) Electron binding energies of aqueous alkaliand halide ions: EUV photoelectron spectroscopyof liquid solutions and combined ab initio andmolecular dynamics calculations. J. Am. Chem.Soc. 127, 7203–7214.

80) Nishizawa, K., Kurahashi, N., Sekiguchi, K., Mizuno,T., Ogi, Y., Horio, T., Oura, M., Kosugi, N. andSuzuki, T. (2011) High-resolution soft X-rayphotoelectron spectroscopy of liquid water. Phys.Chem. Chem. Phys. 13, 413–417.

81) Staib, A. and Borgis, D. (1996) Reaction pathways inthe photodetachment of an electron from aqueouschloride: A quantum molecular dynamics study. J.Chem. Phys. 104, 9027–9039.

82) Hart, E.J. and Boag, J.W. (1962) Absorptionspectrum of hydrated electron in water and inaqueous solutions. J. Am. Chem. Soc. 84, 4090–4095.

83) Bragg, A.E., Verlet, J.R.R., Kammrath, A.,Cheshnovsky, O. and Neumark, D.M. (2004)Hydrated electron dynamics: From clusters tobulk. Science 306, 669–671.

84) Coe, J.V., Lee, G.H., Eaton, J.G., Arnold, S.T.,Sarkas, H.W., Bowen, K.H., Ludewigt, C.,Haberland, H. and Worsnop, D.R. (1990) Photo-electron-spectroscopy of hydrated electron clusteranions, (H2O)!nF2–69. J. Chem. Phys. 92, 3980–3982.

85) Coe, J.V., Earhart, A.D., Cohen, M.H., Hoffman,G.J., Sarkas, H.W. and Bowen, K.H. (1997) Usingcluster studies to approach the electronic structureof bulk water: Reassessing the vacuum level,conduction band edge, and band gap of water. J.Chem. Phys. 107, 6023–6031.

86) Coe, J.V., Williams, S.M. and Bowen, K.H. (2008)Photoelectron spectra of hydrated electron clustersvs. cluster size: connecting to bulk. Int. Rev. Phys.Chem. 27, 27–51.

87) Turi, L., Sheu, W.S. and Rossky, P.J. (2005)Characterization of excess electrons in water-

cluster anions by quantum simulations. Science309, 914–917.

88) Turi, L., Sheu, W.S. and Rossky, P.J. (2005)Response to comment on “Characterization ofexcess electrons in water-cluster anions by quan-tum simulations”. Science 310, 1769.

89) Shkrob, I.A. (2007) The structure of the hydratedelectron. Part 1. Magnetic resonance of internallytrapping water anions: A density functional theorystudy. J. Phys. Chem. A 111, 5223–5231.

90) Shkrob, I.A., Glover, W.J., Larsen, R.E. andSchwartz, B.J. (2007) The structure of thehydrated electron. Part 2. A mixed quantum/classical molecular dynamics embedded clusterdensity functional theory: Single-excitation config-uration interaction study. J. Phys. Chem. A 111,5232–5243.

91) Larsen, R.E., Glover, W.J. and Schwartz, B.J. (2010)Does the hydrated electron occupy a cavity?Science 329, 65–69.

92) Jacobson, L.D. and Herbert, J.M. (2011) Theoreticalcharacterization of four distinct isomer types inhydrated-electron clusters, and proposed assign-ments for photoelectron spectra of water clusteranions. J. Am. Chem. Soc. 133, 19889–19899.

93) Herbert, J.M. and Jacobson, L.D. (2011) Nature’smost squishy ion: The important role of solventpolarization in the description of the hydratedelectron. Int. Rev. Phys. Chem. 30, 1–48.

94) Shreve, A.T., Yen, T.A. and Neumark, D.M. (2010)Photoelectron spectroscopy of hydrated electrons.Chem. Phys. Lett. 493, 216–219.

95) Horio, T., Shen, H., Adachi, S. and Suzuki, T. (2012)Photoelectron spectra of solvated electrons in bulkwater, methanol, and ethanol. Chem. Phys. Lett.535, 12–16.

96) Kammrath, A., Verlet, J.R.R., Griffin, G.B. andNeumark, D.M. (2006) Photoelectron imaging oflarge anionic methanol clusters: (MeOH)n!

(n 9 70–460). J. Chem. Phys. 125, 171102.

(Received June 29, 2012; accepted Oct. 12, 2012)


Profile

Toshinori Suzuki was born in 1961. As a freshman in Tohoku University he wasinspired by quantum chemistry lectures given by the late Prof. Yoshito Amako, and hedecided to study physical chemistry. He graduated from the university in 1984, andreceived the doctoral degree from the same university in 1988 under the supervision ofProf. Mitsuo Ito. He immediately joined the research group of Prof. Eizi Hirota at theInstitute for Molecular Science as a technical associate and later became a researchassociate. During this period, he investigated O(1D) reaction with methane by infrareddiode laser spectroscopy, and he decided to develop a crossed-beam ion imagingexperiment for studying polyatomic reactions. In order to realize it, he resigned Institutefor Molecular Science and joined Cornell University (Prof. Paul Houston) and theUniversity of California, Berkeley (Prof. Yuan T. Lee) as a JSPS fellow for oversea studies. In 1992, he wasappointed again by the Institute for Molecular Science as an associate professor and returned to Japan to start hisresearch group. He initiated a crossed-beam imaging experiment and also realized femtosecond photoelectronimaging. In 2001, He moved to RIKEN to be a chief scientist and extended his research further. The crossed-beamimaging experiment on O(1D) reaction with methane was realized finally in 2008, 20 years after his first planning.He started the research project on chemical reaction dynamics in water using the advanced facilities of SPring-8and SACLA (x-ray free electron laser), which is on going. In 2009, he became a professor of physical chemistry atKyoto University to inspire, and be inspired by, young students. He received Broida Award, IBM Science Award,JSPS Prize and Commendation for Science and Technology by MEXT. He served as a president of Japan Societyfor Molecular Science for 2010–2012.


visualization of chemical reaction dynamics: toward ...€¦ · keywords: molecular beam, chemical...

Documents